Neuronal basis of innate olfactory attraction to ethanol in Drosophila

Abstract

The decision to move towards a mating partner or a food source is essential for life. The mechanisms underlying these behaviors are not well understood. Here, we investigated the role of octopamine - the invertebrate analogue of noradrenaline - in innate olfactory attraction to ethanol. We confirmed that preference is caused via an olfactory stimulus by dissecting the function of the olfactory co-receptor Orco (formally known as OR83b). Orco function is not required for ethanol recognition per se, however it plays a role in context dependent recognition of ethanol. Odor-evoked ethanol preference requires the function of Tbh (Tyramine β hydroxalyse), the rate-limiting enzyme of octopamine synthesis. In addition, neuronal activity in a subset of octopaminergic neurons is necessary for olfactory ethanol preference. Notably, a specific neuronal activation pattern of tyraminergic/octopaminergic neurons elicit preference and is therefore sufficient to induce preference. In contrast, dopamine dependent increase in locomotor activity is not sufficient for olfactory ethanol preference. Consistent with the role of noradrenaline in mammalian drug induced rewards, we provide evidence that in adult Drosophila the octopaminergic neurotransmitter functions as a reinforcer and that the molecular dissection of the innate attraction to ethanol uncovers the basic properties of a response selection system.

Figure 1. Ethanol preference is based on olfactory information. a

Flies are offered a choice between two odor traps and flies will normally decided for one or the other trap within 16 h. The preference index (PI) indicates the percentage of flies that prefer one odor over the other. A positive PI is defined as preference and a negative PI indicates aversion. b Control and heterozygote Orco¹ mutants prefer ethanol containing food odors, whereas transheterozygote Orco¹/Orco² and homozygote Orco¹ mutants do not show preference (PIs are w¹¹¹⁸ 0.19±0.07, Orco¹/+0.55±0.05, Orco¹/Orco² −0.01±0.07 and Orco¹/Orco¹ 0.11±0.08. n = 24, 26, 32 and 31. P<0.01 and P<0.001). c Flies prefer food odors over water (PIs are w¹¹¹⁸ 0.91±0.02 and w¹¹¹⁸; Orco¹ 0.77±0.03. n = 29. There is a significant difference for food preference between w¹¹¹⁸ and Orco¹. P<0.05) and ethanol over water (PIs are w¹¹¹⁸ 0.76±0.06 and w¹¹¹⁸;Orco¹ 0.8±0.04. n 11 and 13.). d Grey bars represent the choice between food odor and ethanol (PIs are w¹¹¹⁸ 0.69±0.04 and w¹¹¹⁸;Orco¹ 0.52±0.05. n 33.) and dark grey indicates the choice between food odor with ethanol versus ethanol alone (PI are w¹¹¹⁸ 0.87±0.03 and w¹¹¹⁸;Orco¹ 0.52±0.06. n = 29 and 30. The olfactory preferences of w¹¹¹⁸ for different tested conditions differ significantly. P<0.05). Bars labeled with a are significantly different from random choice as determined by One-sample sign test.

Figure 2. Ethanol preference depends on TbH function.

a Tbh^nM18 mutants do not show preference and differ significantly from controls (PIs are Tbh^nM18 0.09±0.09 and control 0.23±0.08. n 21 and 23. P<0.01). The loss of preference is independent of w¹¹¹⁸ (PIs are w¹¹¹⁸ 0.31±0.06 and w¹¹¹⁸, Tbh^nM18 0.08±0.09. n 26 and 17. P<0.05 and P<0.01). b Tdc2-GAL4 dependent expression of TbH in Tbh^nM18 mutants restores preference to control levels (PIs are w¹¹¹⁸;Tdc2-GAL4/+0.34±0.05, w¹¹¹⁸; UAS-TbH 0.36±0.07, w¹¹¹⁸,Tbh^nM18;Tdc2-GAL4/+ −0.14±0.08, w¹¹¹⁸,Tbh^nM18;UAS-TbH/+ −0.01±0.08 and w¹¹¹⁸,Tbh^nM18;Tdc2-GAL4/UAS-TbH 0.34±0.1. n 21, 28, 23, 21 and 21.). Mutants carrying the transgenes insertions in the mutant background differ significantly from control carrying the respective transgene (P<0.05 and P<0.01). c Tbh^nM18 mutants prefer food odor over water similar to the controls (PIs are for w¹¹¹⁸ 0.77±0.03 and w¹¹¹⁸, Tbh^nM18 0.81±0.03. n 18 and 15.) but do not prefer ethanol odor over water to the same extend as the controls (PIs are w¹¹¹⁸ 0.67±0.05 and w¹¹¹⁸,Tbh^nM18 0.31±0.08. n 20 and 21. P<0.001). d Control flies prefer complex odors over the single odor ethanol (PI is w¹¹¹⁸ 0.36±0.06. n 35.) and increase their preference significantly for the complex odor when ethanol is added (PI is w¹¹¹⁸ 0.81±0.03. n 18, P<0.001). Tbh^nM18 flies show an increased preference for complex odors with ethanol over ethanol in comparison to the choice of complex odors without ethanol and ethanol (w¹¹¹⁸,Tbh^nM18 0.10±0.08 and 0.74±0.04. n 24 and n 17. P<0.001). The letter a marks differences from random choice as determined by a One-sample sign test.

Figure 3. The Tdc2-GAL4 activates UAS transgenes in a subset of TbH positive neurons.

The GAL4 expression domain of the Tdc2-GAL4line is visualized by GFP in comparison to TbH expression in the anterior (a) and posterior (b) part of the adult brain. Neurons expressing TbH (G0a) or GFP (G5a) and neurons expressing both (VMI) are highlighted with an arrow (a). In the back of the brain G4a neurons only express TbH, whereas VMII and VMIII neurons express both (b). The expression of Tbh (c to f) and GFP (c′ to f′) is found in the same set of neurons of the G3a/AL2 cluster (c′′), VMI (d′′), VMII (e′′) and VMIII (f′). The analysis of expression patterns is summarized in table g and clusters in which GFP and Tbh co-localize are highlighted in grey. The scale bar represents 25 µm.

Figure 4. TbH is required in a subset of octopamingergic neurons for ethanol preference.

a Expression of TbH in a Tdc2-GAL4; Cha-GAL80 dependent manner in Tbh^nM18 mutants does not restore preference (PIs are Tdc2-GAL4/+; Cha-GAL80/+: 0.26±0.05, Tbh^nM18; Tdc2-GAL4; Cha-GAL80/+0.23±0.08 and Tbh^nM18, UAS-Tbh; Tdc2-GAL4; Cha-GAL80/+0.00±0.10, n 27, 22 and 25. The experimental group does not develop preference and differs significantly from the control. The transgene insertion in the mutant background differs significantly from respective control. P<0.05 and <0.001.). b Feb15-GAL4 dependent expression of TbH in Tbh^nM18 mutants restores preference (PIs are w¹¹¹⁸; Feb15-GAL4/+0.37±0.07; w¹¹¹⁸, Tbh^nM18; Feb15-GAL4/+0.14±0.07 and w¹¹¹⁸, Tbh^nM18, UAS-Tbh; Feb15-GAL4/+0,47±0.04. n 28, 33 and 36. The level of preference of the experimental group does not differ from the control. The transgene insertion in the mutant background differs significantly from respective control. P<0.05.). c Loss of preference in Tbh^nM18 mutants is not completely restored by TbH expression under the control of the 6.2-Tbh-GAL4 driver (PIs are w¹¹¹⁸; 6.2-Tbh-GAL4/+0.29±0.06, w¹¹¹⁸, Tbh^nM18; 6.2-Tbh-GAL4/+0.06±0.08 and w¹¹¹⁸, Tbh^nM18, UAS-Tbh; 6.2-Tbh-GAL4/+0.13±0.08. n 36, 26 and 36. The olfactory preferences of the control group and experimental group are not significantly different. Mutants carrying one copy of the GAL4 transgene differ significantly from the control carrying one copy of the GAL4 transgene. P<0.05.). The schemata below graphs for behavioral experiments summarize the expression domains of respective GAL4 drivers in comparison to TbH expression. White cells are TbH positive neurons not targeted by specific driver lines and grey cells indicate co-localization of GAL4 and TbH. The letter a labels an olfactory preference score that is significantly different from random choice as determined by One-sample sign test.

Figure 5. Neuronal activity is required for preference.

After activation of UAS-Kir2.1 under the control of Tdc2-GAL4, tub-GAL80^ts preference is significantly reduced (PIs are Tdc2-GAL4, tub-GAL80^ts; UAS-Kir2.1 0.4±0.1 and with HS 0.12±0.09. n 22. P<0.05). Activation of a UAS-mCD8::GFP construct does not interfere with preference (PIs are Tdc2-GAL4, tub-GAL80^ts; UAS-mCD8::GFP 0.22±0.16 and with HS 0.54±0.17. n 14.). Transgene insertions do not alter preference of flies carrying UAS-Kir2.2 and tub-GAL80^ts (PIs are w¹¹¹⁸; tub-GAL80^ts; UAS-Kir2.1 0.29±0.11 and with HS 0.37±0.09. n 27 and 28.).

Figure 6. Neuronal activity is sufficient to induce preference. a

The behavioral set up of the optogentic trap assay consists of two odor traps filled with food odor surrounded by a dark intransparent plastic. The assay is placed on a cold light plate to motivate flies to descend into the fly traps. On top of the two odor traps two different diodes – one for blue light and one for yellow – are placed that can be activated with the different frequencies. b Light activation of neurons in a Tdc2-GAL4 dependent manner causes preference for the trap illuminated in blue (PIs are norp^A1, Tdc2-GAL4/UAS-ChR2;UAS-ChR2 with vehicle −0.02±0.11 and with retinal 0.35±0.09; n 12 for both. P<0.05.). c Typical traces of norp^A1,Tdc2-GAL4/UAS-ChR2;UAS-ChR2 and norp^A1,TH-GAL4/UAS-ChR2;UAS-ChR2 flies before and after a one min illumination period are shown. The average activity of 7 flies per genotype and condition are summarized in the table. d Activation of dopaminergic neurons causes aversion for the trap illuminated in blue (PIs are norp^A1,TH-GAL4/UAS-ChR2;UAS-ChR2 with vehicle −0.07±0.09 and with retinal −0.41±0.1. n 20 and 35. P<0.05.).